The approaches described in this section are approaches that could be pursued, but not necessarily approaches that have been previously conceived or pursued. Therefore, unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section.
Data is commonly stored in computer-based systems in fixed, rigidly structured data stores. For example, one common type of data store is a “flat” file such as a spreadsheet, plain-text document, or XML document. Another common type of data store is a relational database comprising one or more tables. Other examples of data stores that comprise structured data include, without limitation, files systems, object collections, record collections, arrays, hierarchical trees, linked lists, stacks, and combinations thereof.
Often, the underlying structure of such data stores is poorly suited to data analysis. One approach for facilitating a more efficient analysis of data in such data stores is to reorganize that data according to an object model that defines object structures and relationships between the object structures. To create the object model, data elements in underlying data stores, such as table rows or cells, may be mapped to properties of the objects in the model. The semantics, or “meanings,” of the various components of the object model are defined by an ontology that categorizes objects, relationships, and/or properties according to various defined types. For example, an ontology might categorize objects as being of one of the following types: person, entity, or event. The example ontology might further define different properties for each object type, such as names, dates, locations, documents, media, and so forth. The example ontology might further define relationships between objects, such as employee, participant, sibling, and so forth.
One approach for generating an ontology-based object model is described in U.S. Pat. No. 7,962,495 B2, issued Jun. 14, 2011, the entire contents of which are hereby incorporated by reference for all purposes. The '495 patent describes a dynamic ontology, in which both the object model and the semantics of the ontology may change and evolve over time as needed for analysis.